ORIGINAL ARTICLE
Advanced Nursing Process quality: Comparing the International Classification for Nursing Practice (ICNP) with the NANDAInternational (NANDA-I) and Nursing Interventions Classification (NIC) Eneida Rejane Rabelo-Silva, Ana Carla Dantas Cavalcanti, Maria Cristina Ramos Goulart Caldas, Amalia de Fatima Lucena, Miriam de Abreu Almeida, Graciele Fernanda da Costa Linch, Marcos Barragan da Silva and Maria M€ uller-Staub
Aims and objectives. To assess the quality of the advanced nursing process in nursing documentation in two hospitals. Background. Various standardised terminologies are employed by nurses worldwide, whether for teaching, research or patient care. These systems can improve the quality of nursing records, enable care continuity, consistency in written communication and enhance safety for patients and providers alike. Design. Cross-sectional study. Methods. A total of 138 records from two facilities (69 records from each facility) were analysed, one using the NANDA-International and Nursing Interventions Classification terminology (Centre 1) and one the International Classification for Nursing Practice (Centre 2), by means of the Quality of Diagnoses, Interventions, and Outcomes instrument. Quality of Diagnoses, Interventions, and Outcomes scores range from 0–58 points. Nursing records were dated 2012–2013 for Centre 1 and 2010–2011 for Centre 2. Results. Centre 1 had a Quality of Diagnoses, Interventions, and Outcomes score of 3546 (645), whereas Centre 2 had a Quality of Diagnoses, Interventions, and Outcomes score of 3172 (462) (p < 0001). Centre 2 had higher scores in the ‘Nursing Diagnoses as Process’ dimension, whereas in the ‘Nursing Diagnoses as Product’, ‘Nursing Interventions’ and ‘Nursing Outcomes’ dimensions, Centre
Authors: Eneida Rejane Rabelo-Silva, ScD, RN, Professor, Graduate Program in Nursing at Federal University of Rio Grande do Sul and Hospital de Clınicas de Porto Alegre and Grupo de Estudo e Pesquisa em Enfermagem no Cuidado ao Adulto e Idoso (GEPECADI), Porto Alegre, RS; Ana Carla Dantas Cavalcanti, ScD, RN, Professor, Universidade Federal Fluminense, Niter oi, RJ; Maria Cristina Ramos Goulart Caldas, RN, Nurse of Instituto Nacional do C^ancer (HC III), Rio de Janeiro, RJ; Amalia de Fatima Lucena, ScD, RN, Professor, Graduate Program in Nursing at Federal University of Rio Grande do Sul and Hospital de Clınicas de Porto Alegre and Grupo de Estudo e Pesquisa em Enfermagem no Cuidado ao Adulto e Idoso (GEPECADI), Porto Alegre, RS; Miriam de Abreu Almeida, ScD, RN, Professor, Graduate Program in Nursing at Federal University of Rio Grande do Sul and Hospital de Clınicas de Porto Alegre and Grupo de Estudo e Pesquisa em
© 2016 John Wiley & Sons Ltd Journal of Clinical Nursing, doi: 10.1111/jocn.13387
What does this paper contribute to the wider global clinical community?
Studies comparing the quality of
nursing records as a function of different standardised terminologies have not been published hitherto. From a quality standpoint, lack of consistency influences care continuity, nursing process documentation and patient safety. The outcomes of such investigations can be used to safely define an optimal strategy to make nurses’ work visible and to evaluate nursing care quality and effectiveness.
Enfermagem no Cuidado ao Adulto e Idoso (GEPECADI), Porto Alegre, RS; Graciele Fernanda da Costa Linch, ScD, RN, Professor, Universidade Federal de Ci^encias da Sa ude de Porto Alegre, Porto Alegre, RS; Marcos Barragan da Silva, MSc, RN, PhD Student, Graduate Program in Nursing at Federal University of Rio Grande do Sul and Hospital de Clınicas de Porto Alegre and Grupo de Estudo e Pesquisa em Enfermagem no Cuidado ao Adulto e Idoso (GEPECADI), Porto Alegre, RS, Brazil; Maria M€ uller-Staub, PhD, RN, FEANS, Professor, Nursing Projects, Research and Innovation (Switzerland) & Hanze University, Groningen, the Netherlands Correspondence: Eneida Rejane Rabelo da Silva, Professor, Escola de Enfermagem da Universidade Federal do Rio Grande do Sul, Rua: S~ ao Manoel, 963 – Rio Branco, Porto Alegre, RS 90620-110, Brazil. Telephone: +55 51 33598017/33598657. E-mails:
[email protected];
[email protected]
1
ER Rabelo-Silva et al.
1 exhibited superior performance; acceptable reliability values were obtained for both centres, except for the ‘Nursing Interventions’ domain in Centre 1 and the ‘Nursing Diagnoses as Process’ and ‘Nursing Diagnoses as Product’ domains in Centre 2. Conclusion. The quality of nursing documentation was superior at Centre 1, although both facilities demonstrated moderate scores considering the maximum potential score of 58 points. Reliability analyses showed satisfactory results for both standardised terminologies. Relevance to clinical practice. Nursing leaders should use a validated instrument to investigate the quality of nursing records after implementation of standardised terminologies.
Key words: classification, International Classification for Nursing Practice, NANDA-International, Nursing Interventions Classification, nursing process, nursing records, Quality of Diagnoses, Interventions, and Outcomes instrument Accepted for publication: 29 April 2016
Introduction In response to criticism regarding the lack of content and validity of the traditional nursing process, the socalled advanced nursing process was developed. Instead of using free-text terms unsupported by evidence to formulate nursing diagnoses, the advanced nursing process uses standardised, validated, evidence-based concepts that are defined in specific nursing classifications. In other words, the advanced nursing process consists of defined, evidence-based concepts and includes valid assessment tools and well-defined nursing diagnoses, interventions and outcomes that are standardised and explained in scientifically based nursing classifications (M€ uller-Staub et al. 2015). Any method used to implement the advanced nursing process must be context-appropriate and able to guide clinical practice. Most commonly, this process is taught and implemented by means of standardised nursing languages (Ackley et al. 2008, Ackley & Ladwig 2014). Standardised nursing languages (SNLs) are attempts to create consistency in the vocabulary used by nurses and offer classification-based structures for a variety of elements involved in care practices. SNLs support professional growth and stability by fostering critical thinking as part of the decision-making process for nursing diagnoses, interventions and outcomes. SNLs allow the proper keeping and refinement of nursing documentation as part of the electronic health record (EHR), thus increasing visibility of
2
nursing as a scientific discipline (Jones et al. 2010, Keenan et al. 2013).
Background The use of SNLs has been described as a contributing factor to the quality of nursing, as it enables care continuity through consistency in communication and documentation of information. This, in turn, increases safety for patients and providers alike. Furthermore, SNLs facilitate goal and outcome-oriented planning and evaluations of the advanced nursing process in clinical practice (M€ uller-Staub et al. 2008b, Saranto & Kinnunen 2009, Linch et al. 2010). Another relevant reason to use SNLs is technological advancement and the need for EHR implementation. The use and permanent storage of EHRs, particularly when combined with SNLs, allows to capture, represent, access, communicate and research nursing documentation information (Keenan et al. 2008, 2012, 2013, Jones et al. 2010, Kelley et al. 2011, Brokel et al. 2012). The outcomes achieved by patients are the most important indicator of quality in health care, of which nursing care is an integral part. Health care communicated in unspecific, unclear wording leads to inconsistencies due to divergences in naming patients’ care needs (nursing diagnoses), nursing interventions and treatment goals between providers, teams, sectors or facilities. Unclear or nonprecise documentation and communication leads to missing care effectiveness, and evaluations of the care given are
© 2016 John Wiley & Sons Ltd Journal of Clinical Nursing
Original article
hampered (Koczmara et al. 2005, 2006, Wang et al. 2011, Zegers et al. 2011). SNLs aim to clearly name evidencebased nursing diagnoses, interventions and outcomes and thus allow the development of consented guidelines (H€ ayrinen et al. 2010, Jones et al. 2010, Paans et al. 2010). For these reasons, the use of SNLs in the advanced nursing process has mobilised nurses worldwide to face the challenge of making the elements of clinical nursing practice universal (H€ayrinen et al. 2010, Kelley et al. 2011, Wang et al. 2011). In choosing an SNL including the advanced nursing process, nurses and their employers take a stance towards keeping patient records based on their knowledge and experience. Not choosing the advanced nursing process is a decision that may lead to omission of data that would otherwise play an important stewardship role in nursing, as nursing records that do not follow a systematic methodology can jeopardise the quality of patient care (H€ayrinen et al. 2010, Tastan et al. 2014). Therefore, the development of scientific evidence in the field of nursing requires utilisation of the advanced nursing process based on SNLs, which enable the use of universal, validated concepts and interpretation of nursing records by different providers at different institutions (Keenan et al. 2008, Anderson et al. 2009, Kelley et al. 2011). Since 1970s, investigators have worked to classify nursing phenomena for the purposes of electronic documentation. Worldwide, these researchers assembled in associations and/or groups with the intent of developing and investigating SNLs. Examples include NANDA-International (NANDA-I), the Center for Nursing Classification and Clinical Effectiveness (CNC), the International Council of Nursing (ICN), the Association for Common European Nursing Diagnoses, Interventions and Outcomes (ACENDIO) and the Asociaci on Espa~ nola de Nomenclatura, Taxonomıa y Diagn osticos de Enfermerıa (AENTDE), among others with similar purposes (Tastan et al. 2014). Studies driven by these groups and by other investigators worldwide have strengthened and improved SNLs, which can be understood as systems of labels of the three nursing elements of diagnosis, outcomes and interventions. These systems include NANDA-I (the most used diagnoses classification worldwide) (Tastan et al. 2014), the Nursing Outcomes Classification (NOC) and the Nursing Interventions Classification (NIC), which provide SNL for nursing outcomes and interventions respectively (Head et al. 2011, Kelley et al. 2011, Johnson et al. 2012). Some systems propose a combination of all three elements, such as the Omaha System, the Perioperative Nursing Data Set (PNDS) and the Clinical Care Classification/ Home Health Care Classification (CCC/HHCC) (Tastan © 2016 John Wiley & Sons Ltd Journal of Clinical Nursing
Advanced nursing process quality by the Q-DIO
et al. 2014) or the NANDA-I linked with NIC and NOC (NNN) (Johnson et al. 2012, Ackley & Ladwig 2014). On the other hand, the International Classification for Nursing Practice (ICNP) includes two of three elements: nursing diagnoses and interventions (Coenen et al. 2012). Various SNLs are employed by nurses worldwide, whether for teaching, research and/or patient care (M€ ullerStaub et al. 2015). Therefore, choosing an SNL poses a challenge for nurse leaders when plans for the implementation of the advanced nursing process and EHR documentation have to be made. Up to now, there has been little research on the quality of SNL application in practice. Questions remain as to whether one classification is superior for clinical documentation of the advanced nursing process or which classification best adapts to the characteristics of a facility. The ICNP proposes a wide-ranging, understandable reference terminology that can adapt to multiple purposes in different countries and be regarded as a significant resource for the description of nursing practices. It does not contain defined nursing diagnoses including related factors and defining characteristics; rather, the ICNP develops catalogues (nursing data subsets) for specified health situations that are used in building health information systems (Hardiker & Rector 2001, Coenen et al. 2012, Tastan et al. 2014). However, it currently employs a seven-axis model. Establishing a nursing diagnosis (ND) requires that the clinical nurse select at least one term from two axes, namely, the axes ‘focus’ and ‘judgement’. Focus is the area of relevance to nursing practice, whereas judgement is the clinical opinion related to the selected focus (Comit^e Internacional de Enfermeiros 2011). NANDA-I promotes clinical reasoning and diagnostic precision by describing each ND as a concept including defining characteristics and related factors (Herdman & Kamitsuru 2014). Stating a ND in practice requires identification of the relevant data (defining characteristics and related or risk factors) collected during nursing assessments (interview, physical examination, laboratory and other test results) for an evidence-based diagnostic validation. In NANDA-I, a ND is a concept constructed by means of a multi-axial system consistent with the ISO reference terminology model (NANDA-I 2015). As in the ICNP (Comit^e Internacional de Enfermeiros 2011), the focus and judgement axes play an essential role in stating ND for patients. The NIC is a wide-ranging intervention classification encompassing the entire domain of nursing as a discipline and representing all areas of nursing practice (NANDA-I 2015). Despite the importance of SNLs as a means of improving the quality of nursing records and supporting nursing
3
ER Rabelo-Silva et al.
practice based on critical thinking and EHR development, we found few studies conducted in clinical settings. Evaluations on the two most widely used SNLs to compare the quality of records using one or the other are missing. The relevance of this study is its focus on measuring the quality of the advanced nursing process based on the NANDA-I/ NIC and ICNP classifications in two Brazilian health facilities. The Quality of Diagnoses, Interventions, and Outcomes (Q-DIO) (M€ uller-Staub et al. 2008a,b, M€ uller-Staub et al. 2009, Saranto & Kinnunen 2009), an instrument recently validated for use in Brazil (Linch et al. 2015), allows to bridge this knowledge gap. Hence, the results of this study provide evidence on the quality of nursing records kept with the NANDA-I/NIC and ICNP standardised terminologies. Simultaneously, it provides inputs to support selection of the SNL most likely to ensure consistency in communication and documentation, thus ensuring patient and provider safety and facilitating goal- and outcome-oriented planning of clinical practice.
The study Aims and objectives To assess the quality of the advanced nursing process in nursing documentation in two hospitals, one using the NANDA-I and NIC terminology and the other using ICNP, by means of the Q-DIO instrument.
Design A cross-sectional study was carried out at two Brazilian hospitals.
Sample/Participants Data were collected from the health records of oncology patients with breast neoplasms who had been hospitalised in the clinical or surgical wards of both study facilities for at least four consecutive days. These facilities were selected because one uses EHRs in combination with the NANDA-I and NIC (Centre 1), whereas the other keeps handwritten nursing records based on the ICNP (Centre 2). Both are teaching hospitals with research centres and both are accredited by Joint Commission International (JCI), a network of health facilities recognised for their excellence in patient care. The study included nursing records dated 2012–2013 for Centre 1 and 2010–2011 for Centre 2. Selection of
4
these periods was determined by the timing of the JCI accreditation process at each facility, as both had conducted training activities pertaining to nursing records in the preceding years. Handwritten records that were illegible to the investigators were excluded from the sample. Electronic records that did not correspond to patients with a clinical diagnosis of malignancy for the aforementioned periods and records from off-site patients were excluded. First, the health record numbers of patients admitted to the study centres with malignant neoplasms and otherwise eligible for the study were compiled in a list. This list was then used for randomisation in the Statistical Package for the Social Sciences (SPSS) 20.0, version software (Chicago, Ilinois, USA), taking into account a 20% attrition rate. Two illegible manual records and seven records that did not correspond to a diagnosis of malignancy had to be excluded from the sample. To prevent bias during Q-DIO completion, data collection was carried out at both facilities by the same researcher, who had been previously trained to administer the Brazilian version of the instrument. Initially, we conducted a pilot study consisting of application of the Q-DIO to 10 health records from each centre by two investigators, which yielded a j coefficient of 085. A noninferiority analysis was used to calculate the sample on the basis of a standard deviation of 28 for Centre 1 and 372 for Centre 2. The sample size was estimated with WINPEPI v11.32, version, Jerusalem, Israel (2011). For a statistical power of 90%, a significance level of 25% (or 0025) and a twopoint difference in scores between centres, the sample size was calculated as 138 health records (69 per facility).
Data collection After sample size calculation and randomisation of records, the investigators collected information contained in the nursing records of history and physical examination (completed at the time of patient admission) and assessed the progress notes kept for the four subsequent days. The nursing diagnoses, interventions, prescriptions and outcomes documented during this period in these notes were evaluated. Centre 1 used the SOIC (subjective, objective, interpretation and management) approach as the standard format for daily record-keeping, whereas Centre 2 used the SOAP (subjective, objective, assessment and plan) format. To assess the quality of the advanced nursing process in nursing documentation, we employed the Brazilian version © 2016 John Wiley & Sons Ltd Journal of Clinical Nursing
Original article
tests were used to compare Cronbach’s a coefficients between the facilities. p-Values < 005 were considered significant.
Results Total Q-DIO score A total of 69 health records from each facility were examined for evaluation. Figure 1 shows the mean total Q-DIO score (maximum: 58). Centre 1, which uses NANDA-I/NIC classifications, had a Q-DIO score of 3546 645, whereas Centre 2, which uses the ICNP, had a Q-DIO score of 3172 462 (p < 0001).
Q-DIO dimension scores Table 1 shows significant differences between facilities in QDIO scores across all dimensions. In the ‘Nursing Diagnoses 60 50
Q-DIO score
of the Q-DIO (Linch et al. 2015), following the guideline for Q-DIO application. This questionnaire consists of 29 items subdivided into four scales, each of which is scored on a three-point scale (0 = not achieved; 1 = partially achieved; 2 = fully achieved), for a minimum score of 0 (zero) and a maximum score of 58 points. The first subscale, ‘Nursing Diagnoses as Process’, addresses nursing assessment accuracy and comprises 11 items with a maximum score of 22; the optimal score is 2 points per item. This subscale evaluates issues related to the patient assessment and interview. The second subscale, ‘Nursing Diagnoses as Product’, comprises eight items for a maximum score of 16. This subscale addresses the accuracy of the nursing diagnostic label, definition, defining characteristics and related factors. The third subscale, ‘Nursing Interventions’, comprises three items with a maximum score of 6; again, the optimal score is 2 points per item. This subscale addresses intervention effectiveness by asking if the intervention affects the aetiology of the ND including planning and verification after performance. The fourth subscale, ‘Nursing Outcomes’, comprises seven items for a maximum score of 14 and an optimal score of 2 points per item. This subscale measures the quality of nursing-sensitive patient outcomes, respectively, the achievement nursing goals as described in the progress notes. For assessment of interobserver agreement, 20% of records from each facility were randomly selected for evaluation by a second investigator at Centre 1 and a third investigator at Centre 2. Both investigators were trained by the researcher who originally validated the Q-DIO. Interobserver agreement ensures that an instrument yields similar results when applied by different investigators to evaluate the same records.
Advanced nursing process quality by the Q-DIO
40 30 20 10 0
Center 1
Center 2
Figure 1 Quality of Diagnoses, Interventions, and Outcomes scores from the two centres. *Student’s t-test (p < 0.001).
Ethical considerations The local Research Ethics Committee approved this study, and all investigators signed a data use agreement form.
Data analysis Data were organised and analysed in SPSS 20.0, version. Continuous variables were expressed as means and standard deviation, as appropriate to the data distribution. Q-DIO results were expressed considering the maximum total score of 58. The two facilities were compared by means of Student’s t-test. The j statistic was used for analysis of inter-rater agreement, and the prevalence and bias-adjusted j (PABAK-OS) to verify confidence. Cronbach’s a was applied to analyse reliability. Fisher–Bonett © 2016 John Wiley & Sons Ltd Journal of Clinical Nursing
Table 1 Comparison between scores for Centre 1 (NANDA-International/Nursing Interventions Classification) and Centre 2 (International Classification for Nursing Practice) by Quality of Diagnoses, Interventions, and Outcomes (Q-DIO) dimension Q-DIO dimension (Brazilian version) Nursing diagnoses as process Nursing diagnoses as product Nursing interventions Nursing outcomes
Centre 1
Centre 2
p*
965 398
1109 268
0015
1309 216
913 117
07), except for the ‘Nursing Interventions’ domain in Centre 1 and the ‘Nursing Diagnoses as Process’ and ‘Nursing Diagnoses as Product’ domains in Centre 2.
Discussion This study presents the first evidence for the quality of nursing records kept using the NANDA-I/NIC and ICNP standardised terminologies at two different hospitals. According to the results, records at the facility that employs the NANDA-I/NIC classifications demonstrated statistically superior results regarding application of the advanced nursing process. Both facilities demonstrated moderate scores considering the maximum potential score of 58 points. Although the Q-DIO does not provide a hierarchical scale for classification of total scores, both facilities demonstrated moderate scores (Centre 1, 3546; Centre 2, 3172). This reveals some difficulty in use of the NANDA-I/NIC and ICNP standardised terminologies by Table 2 Comparison of Cronbach’s a coefficients between the NANDA International (NANDA-I)/Nursing Interventions Classification (NIC) and International Classification for Nursing Practice (ICNP) terminologies by Quality of Diagnoses, Interventions, and Outcomes (Q-DIO) domain (Brazilian version)
Q-DIO domains Nursing diagnosis (ND) as process ND as product Nursing interventions Nursing outcomes
6
Centre 1 NANDA-I, NIC
Centre 2 ICNP
0762
0519
0003
0834 0599 0847
0532 0799 0810
